256 Mr. S. Bidwell's Experiments illustrating 



be strained when placed between the poles of an electro- 

 magnet. 



My first idea was that, since the magnetic force was stron- 

 gest in the middle of the field, the lower half of the metallic 

 sheet would be, on the whole, transversely compressed and the 

 upper half extended. But I immediately saw that strains of 

 this nature would not explain Hall's phenomenon, unless 

 indeed it turned out to be the case that lateral compression 

 and extension produced opposite thermoelectric effects to those 

 caused by longitudinal compression and extension. The effect 

 of lateral compression in an iron wire was noticed by Thomson 

 in the paper referred to. He states that when an iron wire is 

 laterally compressed, a current passes from uncompressed to 

 compressed through the hot junction. He supposes therefore 

 that since a lateral traction would probably produce the reverse 

 effect of a lateral pressure, it w^ould give rise to currents from 

 stretched to free — the reverse of longitudinal traction. In 

 making this experiment, Thomson used many turns of wire 

 wound round a block of w r ood (to multiply the effect) and 

 squeezed the whole in a hydraulic press. I attempted to repeat 

 the experiment by a simple method and using only one wire; 

 but no amount of pressure that I was able to produce resulted 

 in any indication whatever of a current. I have since endea- 

 voured to ascertain whether any appreciable current could be 

 produced in a sheet of metal by lateral traction. The arrange- 

 ment used is shown in fig. 2: — AB is a sheet of metallic foil; 

 the shaded part D B was held between two fixed blocks of 

 wood, and the shaded part C F between two blocks of wood 

 which were attached to a lever. A heated glass rod was 

 applied along C D, and on drawing the lever upwards in the 

 direction of the arrow, a current immediately passed between 

 the unstrained and the strained portions of the metal through 

 the hot junction. Two metals only have been thus treated — 

 zinc and iron ; and the direction of the current was in both 

 cases the same as if the traction had been longitudinal instead 

 of lateral. In zinc the effect was very strong, in iron it was 

 feeble, but not, I think, so feeble as to be doubtful. 



On further consideration I came to the conclusion that no 

 appreciable strains of the kind above referred to would be 

 produced by the transverse force in the magnetic field. This 

 force would be sensibly uniform across the whole of the 

 middle portion of the strip, for the pole-pieces were flat and 

 within a very short distance of each other, and the strip did 

 not extend across their whole width. Considering the analogy 

 of a beam of wood, rigidly fixed at both ends, and acted upon 

 in the middle by a force perpendicular to its length, I imagined 



